fixed bug in app.py

app.py

@@ -4,8 +4,10 @@ from tokenizers import Tokenizer
 import gradio as gr
 from model import LlamaForCausalLM  # Import your custom model class
 
-# Load your custom tokenizer from tokenizer.json
-tokenizer = Tokenizer.from_file("vocab.json")
+# Load tokenizer and model
+tokenizer = AutoTokenizer.from_pretrained("HuggingFaceTB/cosmo2-tokenizer")
+if tokenizer.pad_token is None:
+    tokenizer.pad_token = tokenizer.eos_token if tokenizer.eos_token else "[PAD]"
 
 # Load the model from a local .pth file
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")